Abstract :
[en] In Tunisia, the dairy sector is currently dealing with a number of new challenges, mainly due to
increasing costs of milk production. Milk production is important because of constantly rising demand. The
production of a cow varies along the lactation, following a curve that can be characterised by different
traits, i.e., the level of production and persistency. Worldwide, by importing international genetic material,
the intensive selection for milk production and components aiming to close the gap between demand and
production has compromised functional traits, i.e., fertility and consequently longevity. Longevity is
important because it can be considered to combine all the characteristics that are directly associated with
the ability of a cow to remain in the herd. To minimise the negative effect of selecting for milk yield on
longevity, studying the relationships between lactation curve traits and longevity seemed of great interest
and was the overall aim of this thesis.
To reach this aim, Chapter 1 presents an overview of the definitions of (1) the lactation curve and its traits,
the models used and the factors of variation, (2) the appreciation of the concept of longevity, the models
used and the factors of variation, and (3) the relationships between longevity and lactation curve traits.
Chapter 2 focused on the study of phenotypic performances of milk production, reproduction and longevity
traits (i.e., lactation survival and lifetime traits) and their relationships. Then, Chapter 3 reports the
estimation of genetic parameters of lactation survival within and across lactation using a multiple-trait
linear random regression model. Finally, the genetic relationships between lactation survival and the
lactation curve traits of milk yield and its components (represented as level and persistency) have been
investigated in Chapter 4 and Chapter 5.
Results showed that cows in Tunisia are not producing up to their potential and that their stay in the herd is
shorter than expected. Defining longevity as a stayability trait (lactation survival) allowed the production of
a multiple-trait model and the computation of correlations with lactation curve traits, i.e., persistency and
level of production. Results showed that lactation survival was positively correlated with total milk yield
and negatively correlated with average components at the genetic level. Persistency was negatively
correlated to lactation survival at the genetic level, suggesting that an efficient selection index should
include both persistency and lactation survival.
The approach used may help clear the way for multiple-trait analysis of lactation survival and lactation
curve traits (e.g., persistency) in order to create sustainable and reliable breeding strategies. Short-term
strategies should focus on improving the performance of recording, both quantitatively and qualitatively. In
a long-term perspective, genomic selection could be interesting for all the studied traits, and especially for
lactation survival given its low heritability, allowing faster genetic progress towards the breeding of dairy
cows adapted to Tunisian conditions.
